The recent description of interstitial fibrosis and renal failure in a streptozotocin model of diabetes in rats subjected to a brief period of renal ischemia provides a good model of human DN and may shed light on the potential pathogenetic mechanisms of this phenomenon. Such new observations, unexplained by the currently prevailing hyperfiltration hypothesis, call for alterative hypoxia-inducible mechanisms of progression in DN. Indeed, it has been demonstrated that hypoxia is itself a potent regulator of gene expression, acting via transcriptional control and mRNA stability to alter the expression of a wide variety of hormones, growth factors, vasoactive compounds and molecules involved in intermediary metabolism. Prominent among hypoxia-inducible growth factors is vascular endothelial growth factor/vascular permeability factor (VEGF/VPF). We hypothesize that the combination of elevated VEGF production, increased post-glomerular vascular permeability and transduction of glycosylated proteins trigger the cascade of events which lead to interstitial fibrosis. Specifically, transudated glycated serum constituents, on the other hand, stimulate fibroblast proliferation, differentiation into myofibroblasts and synthesis of matrix proteins, while on the other hand, these glycosylated products inhibit angiogenesis, thus maintaining the state of chronic hypoxia. This hypothesis will be tested functionally (hemodynamic parameters, tissue oxygenation, expression of angiogenic and angiostatic factors, effects of anti-VEGF and VEGF on these parameters), morphologically (mapping of glycated proteins, proteoglycans, glycoproteins, and other markets of vascular permeability), using approaches of cell biology (isolation of renal fibroblasts, co- cultures of endothelial cells with fibroblasts, parameters characterizing cell cycle and their modification by glycated matrix/serum proteins utilized as a substratum, expression of genes and gene products participating in matrix synthesis and degradation and the influence of glycated proteins on these parameters,) cellular physiological approaches (endothelial cell migration and angiogenic potential under the conditions of dysfunctional NO synthase or perturbed repertoire of glycated matrix proteins, balance between the vascular permeability and angiogenesis as affected by glycated proteins and VEGF, and the potential resolution of fibrosis by angiogenic promoters). The proof of this unifying hypothesis of endothelium-dependent fibroblast activation feeding back to inhibit angiogenesis, will require studies on several levels of complexity-from molecular and cellular biology to whole animal physiology-and warrants the combined efforts of three problem-targeted investigative groups. If proven to be correct, this hypothesis could delineate new therapeutic approaches to prevent the progression of diabetic nephropathy.